Product Metrics & KPIs: The Ones That Actually Matter
AARRR funnel, DAU/MAU, leading vs lagging indicators: how to pick 5-7 product KPIs that drive real decisions and connect them to your OKRs and roadmap.
By the RoadmapHero team · Updated on · 6 min read
Key takeaways
- The right product metrics tell you whether the product is creating more value this month than last month, and where it is leaking value.
- The AARRR framework organizes product KPIs by funnel stage (Acquisition, Activation, Retention, Revenue, Referral), and the funnel should be fixed from the bottom up.
- Steer with 5–7 KPIs, one per AARRR stage plus an engagement ratio and, in B2B, net revenue retention, each with an owner and a target.
- Pair leading indicators such as activation rate with lagging ones such as churn or ARR, ban vanity metrics like cumulative signups, and link KPIs to OKRs and roadmap bets.
Most teams don't suffer from a lack of data — they suffer from a lack of selection. The right product metrics answer one question: is the product creating more value this month than last month, and where is it leaking? This guide walks through the KPIs that matter at each stage of the funnel, the engagement ratios worth watching, the leading/lagging distinction, and how to wire the whole thing into your OKRs and roadmap.
Why most product dashboards fail
The typical failure mode isn't missing data. It's a dashboard with 40 charts where nobody can say which three numbers the team is accountable for. Symptoms: metrics reviews that end without decisions, KPIs that never change what gets built, and a different "key" metric quoted in every meeting. The fix is structural: a small set of chosen product KPIs, each with an owner, a target and a direct line to the roadmap.
AARRR: metrics by funnel stage
The AARRR framework (Dave McClure's "pirate metrics") remains the cleanest way to organize product KPIs, because every metric gets a home and every leak gets a stage:
Stage | Question | Typical metrics |
|---|---|---|
Acquisition | Are the right people finding us? | Signups, cost per acquisition, traffic-to-signup rate |
Activation | Do they reach first value? | Activation rate, time-to-value, onboarding completion |
Retention | Do they come back? | D7/D30 retention, cohort curves, churn rate |
Revenue | Do they pay (and pay more)? | Conversion to paid, MRR/ARR, net revenue retention, ARPA |
Referral | Do they bring others? | NPS, referral rate, viral coefficient (k-factor) |
Two practical rules. First, fix the funnel from the bottom up: pouring acquisition into a leaky retention bucket wastes money, so a startup with 30% D30 churn should not be optimizing ad spend. Second, define activation precisely — the specific action that correlates with long-term retention (e.g. "created a roadmap and invited one teammate within 3 days"), not a vague "finished onboarding".
Engagement: DAU/MAU and stickiness
Between activation and retention lives engagement — how deeply the product is woven into users' routines:
DAU/MAU (stickiness): the share of monthly users who show up on a given day. A DAU/MAU of 25% means the average user is active about 7–8 days a month.
Feature adoption: % of active users who used a given feature in the period. Essential for judging whether last quarter's bets landed.
Depth of use: actions per session, number of workspace members, integrations connected — pick the one that matches your value story.
Benchmark against your product's natural frequency. A payroll tool used twice a month with 100% task completion is healthier than a social app with daily opens and no value delivered. Judging a weekly-use B2B product on DAU/MAU is self-inflicted anxiety.
Leading vs lagging indicators
Lagging indicators (churn, ARR, NPS) tell you what already happened — you can't manage them directly. Leading indicators (activation rate, weekly active teams, support contact rate) move first and predict the lagging ones. A healthy KPI set pairs them: for example, "% of new accounts activated in 7 days" (leading) feeding "net revenue retention" (lagging). When a lagging metric is your goal, always ask: which upstream behavior, measurable this week, drives it? That upstream behavior is what the roadmap can actually act on — the same logic behind a good North Star Metric and its input tree.
Choose 5–7 KPIs, not 40
A selection method that takes one workshop:
Write the value sentence: "our product creates value when a user does X, Y often". Metrics that don't relate to that sentence are out.
Pick one KPI per AARRR stage — five so far.
Add one engagement ratio (DAU/MAU or feature adoption) and, if you're B2B, one revenue-quality metric (net revenue retention).
Assign an owner and a target to each. A KPI without an owner is a chart.
Demote everything else to a diagnostic dashboard you open only when a KPI moves.
Example for a B2B SaaS: signup rate, 7-day activation, weekly active teams, D90 logo retention, net revenue retention, NPS, plus time-to-value as a watch-item. Seven numbers, one screen, zero ambiguity.
Vanity metrics to ban
A vanity metric is any number that can grow while the business gets worse. Classic offenders:
Cumulative anything — total signups, total downloads. Monotonic curves flatter and inform nothing.
Registered users instead of active users.
Page views / sessions without a link to value events.
Features shipped per quarter — output is not outcome (the core argument for an outcome-based roadmap).
Averages that hide distributions — average session time mixing 5-second bounces with 2-hour power sessions.
The test: "if this number doubled, would we know what caused it and would we do anything differently?" If either answer is no, it's vanity.
From metrics to OKRs to roadmap
Metrics only pay off when they close the loop into planning. The chain looks like this:
KPIs reveal the gap: activation is 22%, target is 35%.
OKRs frame the ambition: "New users reach first value fast — activation from 22% to 35%" becomes a Key Result. Our OKR guide for product teams covers how to write these well.
The roadmap carries the bets: onboarding checklist, templates, aha-moment redesign — each initiative linked to the KR it serves, as detailed in aligning your roadmap with OKRs.
Shipped work moves the KPI (or doesn't) — and that result reprioritizes the next cycle.
Teams that skip step 2 get metric-chasing without strategy; teams that skip step 4 get a roadmap that never learns. And when retention is the weak stage, feed the loop with qualitative signal too — our guide on reducing churn with product feedback shows how numbers and verbatims combine.
Tracking this in RoadmapHero
RoadmapHero doesn't replace your analytics stack — it's where metrics become decisions. You define OKRs with measurable Key Results (your chosen KPIs), link backlog items and epics to them, and the HeroScore weighs each item's business value by its contribution to those objectives. The AI progression analysis summarizes how each Key Result is trending and which initiatives are moving it, so your metrics → OKR → roadmap chain lives in one place. Free trial, transparent pricing in euros, no commitment.
Frequently asked questions
How many product KPIs should a team track?
Five to seven for active steering, plus a handful of diagnostic metrics you check when something moves. Beyond that, dashboards stop driving decisions: if a metric wouldn't change what you build next quarter, it doesn't belong on the main board.
What is a good DAU/MAU ratio?
It depends entirely on the natural frequency of your product. Messaging apps can exceed 50%, social products 20–30%, while excellent B2B tools used weekly may sit at 10–15%. Compare against your own history and product category, not a universal benchmark.
What is the difference between a metric and a KPI?
A metric is anything you can measure; a KPI is a metric you have committed to steer by, usually with a target. All KPIs are metrics, but only a few metrics deserve KPI status — that selection is the actual strategic act.
Are vanity metrics ever useful?
They can be useful for storytelling — investor decks, press releases — but never for decisions. Cumulative signups or app downloads only go up, so they can't tell you whether the product is getting better or worse. Keep them out of the team dashboard.
Should revenue be a product KPI?
Revenue belongs on the executive dashboard, but it's a lagging indicator: by the time it moves, the causes are months old. Product teams get more leverage from leading indicators like activation rate, retention cohorts and feature adoption that predict future revenue.